User blog:TenshiAkari12/IA TypeC - 1st Day's Verdict...
After playing around with the alpha VB demo for several hours today... I find there are already a good number of plus points, as well as annoyances. They may seem to be more cons than anything listed here, but hey, this thing is just a test run... it's to be expected they crop up one at a time when they feel like it. :-P Like I mentioned in the description of the demo video I posted(*), she's got a lot more force coming out of her, which can be a blessing or a curse depending on how what you would plan on using this bank for. That, and I'm seeing a lot of people don't hear the difference unless the two banks are played alongside each other or they check out the demo in the V3 Editor... that's the oddest thing about this bank: depending on what songs you listen to, it can either blend in or stick out. It's just going to take some getting used to hearing it in order for fans who aren't familiar with using Vocaloids to pick it out, as per usual for anything new in this fandom. :-D (*) Link is in my other blog post, or you can search for "IA α TypeC" on YouTube and find it with (~TA12~) tacked to the end of it. Pros: *Strong, clear, tone overall. *Works and responds pretty well with positive GEN parameters for the most part. Only one issue noted below with concerns to that, but everything else sounded pretty solid. *Also very responsive with BRI and fairly so with OPN parameters (which is great, because some notes will need tweaking with these two at times). I also love the fact that with a lower brightness setting, her tone is the same as the original when you find the right value for it. It's definitely going to help when it comes to mixing in the different pronunciations of the phonemes between the two. *Still retains most of the recognizable quirks IA's original bank has well (good with long tones, vibratos, phoneme enunciation, etc.) Cons: *Her "あa" phonemes... my GOD, they're out there! They have a tendency to stick out when used in short-valued notes and left unedited, and the result is spready-sounding as all get down, the complete opposite of the original bank where it's one of her highlight phonemes. It's an issue mostly when transitioning notes in somewhat upbeat songs, and is sometimes a mortal enemy of grace notes, the vowel extender, and other hidden phonemes depending on the note's length. (I used my VSQx of BREEZE to see how it would sound, and it's not pleasing in the least... in fact, in half of that one song unedited she sounds like female Ryuut---(*///shot*) ... anyways, in most cases like this, you'd have to make use of the Opening parameter to tone it down some, because it sticks out if not edited. However, it'd be nice if it were the opposite again, instead of having to worry about toning something down have it so you could easily bring it up. *Found a few odd sounding instances with other phonemes, mainly with "うM"... it bends upward a bit and the tone "spreads" at the end of the note on it's own, but it's fine in transitions. That, and when it's used on F#3, there's a strange warble if you happen to have vibrato set on that particular note with the GEN parameter is 20 or greater. * E-flat3 & E3 "のo" phoneme combo sounds a bit "buzzy" and unclear at times, especially with non-vibrato. This seems to be the same issue that's already present in the original bank, and just like that one it's overrideable with edits to the phonemes... though I would like to see it actually adjusted so it sounds decent without the edits. * Glottal stops (?) sometimes stop the next note from ever playing instead of adding a pause in between phonemes. For example:( i ? a i t) is one of the combos I used in my BREEZE cover to mimick the word "delight". With the original VB, it works fine with no such issues. However when TypeC tries to hit it, only the first instance is successful, while with the other three instances, a i is nothing but silence. Strange (both the facts that this is happening and I'm using glottal stops in odd places to substitute for Sil and Asp)... but I could see it possibly being a slight issue if this happens in other random instances with different phonemes. The issue in the first bullet point of the cons? Short-valued or fast notes? Yeah... don't expect this bank to have the same record-breaking BPM as the original, with the change in tone quality, something had to be sacrificed, hopefully just for the time being. It's likely yet another "temporary" issue, though, because we all know it'd be a bad deal to have this type of tone transition issue persist, seeing how popular songs with fast, polyrhythmic lyrics are. So... that's pretty much it today, though there are plenty of things I still think can be experimented with during the time period. Most of the issues can be "overridden" through means of separating phonemes into different notes and adjusting the parameters, but overall, the new bank seems a little harder to work with as it is right now than the original. Hoping they're able to at least balance them out before the full release. Until then, I'm hoping to get a better feel for this bank by the time the trial period ends. By then I'll definitely backtrack and add any other things to note with it. Category:Blog posts